Re: [PATCH RFC v2 2/2] linux/vnuma: Enables NUMA for domu PV guest

[David Vrabel <david.vrabel@citrix.com>]

On 17/09/13 09:34, Elena Ufimtseva wrote:
> After the NUMA topology was received from Xen,
> enable NUMA during boot. Should have CONFIG_NUMA
> enabled in kernel.
> 
> Changes since v1:
> - added additional checks for PV guest and hypercall
> support before enablinf NUMA;

As I said in response to the other patch, I don't think this does the
right thing.

I think xen_vnuma_support() needs to try the vnuma hypercall and check
it is successful.

> --- a/arch/x86/xen/setup.c
> +++ b/arch/x86/xen/setup.c
> @@ -20,6 +20,7 @@
>  #include <asm/numa.h>
>  #include <asm/xen/hypervisor.h>
>  #include <asm/xen/hypercall.h>
> +#include <asm/xen/vnuma.h>
>  
>  #include <xen/xen.h>
>  #include <xen/page.h>
> @@ -583,6 +584,9 @@ void __init xen_arch_setup(void)
>  	WARN_ON(xen_set_default_idle());
>  	fiddle_vdso();
>  #ifdef CONFIG_NUMA
> -	numa_off = 1;
> +	if (!xen_initial_domain() && xen_vnuma_support())

I don't think there's a need to special case the initial domain here is
there?

> +		numa_off = 0;
> +	else
> +		numa_off = 1;
>  #endif
>  }

David